Abstract—Due to wide adoption of smart phones, Location
Based Services (LBS) that leverage the user’s location
information have recently attracted an increasing amount of
interest. As one of the most promising LBS technology, indoor
Wi-Fi based positioning can provide relatively precise location
information of Wi-Fi enabled mobile users at a low cost. In this
paper, we propose a model called “Mixed State-Weighted
Markov-chain Model” (MSWMM) to predict the next location
of a user given his previous n locations, where the locations are
derived from Wi-Fi based positioning. MSWMM is an improved
version of the “Mixed Markov-chain Model” (MMM) and it
takes into account the visited frequency of the same location and
not just the transition probability between adjacent locations. In
the experiment of comparing with MMM for n=2, MSWMM
yields a significant 20.38% improvement of prediction accuracy
over MMM.
Index Terms—Markov-chain model, next location prediction,
Wi-Fi based positioning.
Jian Huang, Boon-Khai Ang, Mun-Lie Seeto, and Hendy Shi are with the
National University of Singapore, 25 Heng Mui Keng Terrace, 119615
Singapore (e-mail: A0092599@nus.edu.sg, A0092597@nus.edu.sg,
A0092671@nus.edu.sg, A0092709@nus.edu.sg).
Daniel Dahlmeier is with SAP Asia Pte Ltd, CREATE Tower 1 Create
Way, 138602 Singapore (e-mail: d.dahlmeier@sap.com).
Ziheng Lin is with Singapore Press Holdings, 1000 Toa Payoh North,
318994 Singapore (e-mail: linziheng@gmail.com).
Cite: Jian Huang, Daniel Dahlmeier, Ziheng Lin, Boon-Khai Ang, Mun-Lie Seeto, and Hendy Shi, "Wi-Fi Based Indoor Next Location Prediction Using Mixed State-Weighted Markov-Chain Model," International Journal of Machine Learning and Computing vol. 4, no. 6, pp. 505-509, 2014.